📝 Sequence definition and examples (70 MCQs)

📖 Explanation: This question targets error analysis regarding the relationship between function limits and sequence limits. A common misconception is confusing the implication direction: if $\lim_{x\to\infty} f(x) = L$, then $\lim_{n\to\infty} f(n) = L$, but the reverse is not true. However, for $\sin(n)$, the sequence does not converge at all. The values of $\sin(n)$ for integer $n$ are dense in $[-1, 1]$ because $\pi$ is irrational, meaning the sequence oscillates indefinitely without settling near 0 or any other limit. Option C correctly identifies the logical flaw and the specific behavior of this sequence.

📖 Explanation: This problem requires conceptual understanding of recursively defined sequences and the Monotone Convergence Theorem. This specific recursion is Newton's Method for finding square roots. For $a_1 > \sqrt{5}$, one can prove algebraically that $a_{n+1} < a_n$ and $a_{n+1} \geq \sqrt{5}$. Since it is decreasing and bounded below, it must converge to a limit $L$. Taking the limit of both sides of the recurrence gives $L = \frac{1}{2}(L + 5/L)$, yielding $L = \sqrt{5}$. Option B captures these necessary conditions precisely, while others describe incorrect behaviors like oscillation or wrong bounds.

📖 Explanation: This is a direct application of the theorem linking function limits to sequence limits. If $\lim_{x \to \infty} f(x) = L$, then for any sequence of inputs going to infinity (like integers $n$), $\lim_{n \to \infty} f(n) = L$. The existence of the horizontal asymptote at $y=3$ for the continuous function guarantees the sequence converges to 3. While the sequence might not be monotone (the function could wiggle while approaching the asymptote), convergence is assured. This distinguishes the forward implication (function to sequence) from the reverse, addressing a key conceptual distinction in analyzing graphs versus discrete data.

📖 Explanation: This question addresses error analysis in applying calculus tools to sequences. L'Hôpital's Rule requires differentiable functions on an interval. The factorial function $n!$ is discrete; while the Gamma function extends it, standard calculus courses treat $n!$ as undefined for non-integers in this context. Even if extended, direct differentiation is impractical. The correct pedagogical approach uses the Squeezing Theorem. By expanding $\frac{n!}{n^n} = \frac{1}{n} \cdot \frac{2}{n} \cdots \frac{n}{n}$, we see it is less than $\frac{1}{n}$ (since all other factors are $\leq 1$). As $1/n \to 0$, the sequence converges to 0. This reinforces choosing appropriate discrete tools over blind continuous ones.

📖 Explanation: This tests the Comparison Test logic for divergence. If a smaller sequence grows without bound, a larger sequence has no choice but to also grow without bound. Formally, since $\lim_{n \to \infty} a_n = +\infty$, for any $M > 0$, there exists $N$ such that $a_n > M$ for $n \geq N$. Since $b_n \geq a_n$, it follows that $b_n > M$ for $n \geq N$, satisfying the definition of divergence to $+\infty$. This is a fundamental property of order and limits. Distractors suggesting possible convergence violate the transitive property of inequalities in the extended real number system.

📖 Explanation: This addresses the classic $1^\infty$ indeterminate form misconception. Students often incorrectly assume $\lim (f(n))^{g(n)} = (\lim f(n))^{\lim g(n)}$. This operation is invalid when the result is indeterminate. The sequence $(1+1/n)^n$ is the definition of Euler's number $e$. Conceptually, the base approaches 1 slowly enough relative to the growth of $n$ that the compound effect accumulates to a finite value greater than 1. Rigorous analysis shows it is increasing and bounded above by 3 (or via binomial expansion), guaranteeing convergence to $e$. Option D is tricky but incorrect because $\ln(\text{limit}) = 1$ implies limit $= e$, not 1.

📖 Explanation: This scenario-based question connects abstract theory to mathematical modeling. The Monotone Convergence Theorem states that a bounded, monotone sequence converges. Chaotic systems like the logistic map at $r=4$ are characterized by sensitive dependence on initial conditions and lack of periodicity or monotonicity. They are bounded (within [0,1]) but definitely not monotone. Therefore, the sufficient condition for convergence provided by monotonicity is absent. This doesn't prove divergence (though chaos implies non-convergence to a fixed point), but it explains *why* that specific tool fails. Option A is irrelevant (recursive definitions are valid); C is factually wrong (it's bounded); D is wrong (the map is continuous).

📖 Explanation: This is an error analysis question focusing on the interaction between magnitude limits and sign oscillation. Convergence requires terms to get arbitrarily close to a *single* number $L$. Here, even-numbered terms approach +1 and odd-numbered terms approach -1. Since $+1 \neq -1$, the limit does not exist. The student correctly computed the limit of the absolute value (magnitude) but erroneously assumed this dictates the limit of the signed sequence. This is a critical distinction: $|a_n| \to L$ does not imply $a_n \to L$ unless $L=0$. The sequence diverges by oscillation, distinct from divergence to infinity.

📖 Explanation: This mixed-concept question contrasts monotone convergence with convergence via squeezing/absolute values. $1/n$ is strictly decreasing to 0. $\sin(n)/n$ oscillates due to the sine term, so it is not monotone. However, the inequality $-1/n \leq \sin(n)/n \leq 1/n$ holds. Since both bounding sequences go to 0, the Squeezing Theorem forces $b_n \to 0$. Alternatively, Theorem 9.1.6 states that if $|a_n| \to 0$, then $a_n \to 0$. Here $|b_n| \leq 1/n \to 0$. This highlights that oscillation does not prevent convergence if the amplitude decays sufficiently, distinguishing rate/bound arguments from monotonicity arguments.

📖 Explanation: This targets a pervasive logical fallacy in handling recursive sequences. Solving $L = f(L)$ finds *candidate* limits (fixed points), but it presupposes that a limit exists. If the sequence were divergent or oscillatory, the equation $L = \sqrt{2+L}$ would still yield $L=2$, but that value would be meaningless as a limit. Rigorous proof requires first establishing convergence, typically via the Monotone Convergence Theorem (showing it's increasing and bounded above by 2). Only after proving $\lim a_n$ exists can one validly apply limit laws to the recurrence. This question enforces the distinction between finding fixed points and proving convergence.

📖 Explanation: This question requires interpreting visual data of sequence behavior. Clustering around two distinct values suggests the sequence does not settle at a single limit. According to Theorem 9.1.4, a sequence converges to $L$ if and only if both its even and odd subsequences converge to $L$. If they converge to different values (1 and -1), the main sequence diverges. This visual pattern is characteristic of sequences like $(-1)^n$ or similar oscillatory forms. It tests the ability to translate graphical clustering into the formal language of subsequence convergence and divergence, ruling out options suggesting convergence to an average or mandatory convergence due to boundedness.

📖 Explanation: This applies sequence limit properties to a realistic modeling variation. We know $\lim (1+r/n)^n = e^r$. For $B_n$, we can split the expression: $(1+r/n)^{n+1} = (1+r/n)^n \times (1+r/n)$. Using the product rule for limits, $\lim B_n = [\lim (1+r/n)^n] \times [\lim (1+r/n)] = e^r \times 1 = e^r$. Thus, both models yield the same continuous compounding limit. This tests understanding that adding a constant or vanishing perturbation to the exponent doesn't change the exponential limit class, reinforcing algebraic manipulation of limits in applied contexts.

📖 Explanation: This tests fundamental theoretical distinctions. While convergence implies boundedness (true), the converse is false. The classic counterexample is $a_n = (-1)^n$, which is bounded between -1 and 1 but diverges by oscillation. Boundedness is a necessary but not sufficient condition for convergence. Monotonicity plus boundedness is sufficient (Monotone Convergence Theorem). Unboundedness implies divergence (contrapositive of 'convergent implies bounded'). Identifying the false statement ensures students don't conflate necessary and sufficient conditions, a common source of error in proof construction and theoretical reasoning.

📖 Explanation: This integrates sequence limits with series convergence, targeting the Divergence Test misconception. The sequence $a_n \to 0$ is verified by rationalization: $\frac{1}{\sqrt{n+1}+\sqrt{n}} \to 0$. However, for the series, partial sums telescope: $S_N = \sqrt{N+1} - \sqrt{1}$. As $N \to \infty$, $S_N \to \infty$. Thus, despite terms vanishing, the sum diverges. This perfectly illustrates that $\lim a_n = 0$ does NOT imply $\sum a_n$ converges. Students must distinguish between the behavior of individual terms (sequence) and cumulative sums (series), avoiding the trap of assuming term decay guarantees summability.

📖 Explanation: This scenario models numerical analysis realities. If $t_n \to \pi$, then for large $n$, $t_n$ is near $\pi$. But $c_n$ is always within $\epsilon = 10^{-6}$ of $t_n$. By triangle inequality, $|c_n - \pi| \leq |c_n - t_n| + |t_n - \pi| < 10^{-6} + |t_n - \pi|$. As $n \to \infty$, $|t_n - \pi| \to 0$, so $c_n$ stays within a $10^{-6}$ neighborhood of $\pi$. It does not necessarily converge to $\pi$ exactly because the error bound is fixed, not vanishing. This highlights the difference between theoretical limits and practical numerical approximations with systematic or rounding errors.

📖 Explanation: Dividing by the wrong power obscures the asymptotic behavior. Dividing by $n^2$ yields $\frac{1+3/n}{2-5/n^2} \to 1/2$, revealing that for large $n$, $a_n \approx \frac{n^2}{2n^2} = 1/2$. Dividing by $n$ leaves dominant terms growing, maintaining indeterminacy. The conceptual error is failing to normalize by the *fastest* growing component to isolate the finite limiting ratio. This reflects a deeper misunderstanding of asymptotic dominance: lower-order terms become negligible only when scaled against the highest order. Correct normalization transforms indeterminate forms into determinate ones by exposing the leading coefficient ratio.

📖 Explanation: Strictly decreasing positive sequences are bounded below by 0. By the Monotone Convergence Theorem, they MUST converge to some limit $L \geq 0$. However, the limit is not necessarily 0. Counterexample: $a_n = 1 + 1/n$ has ratio $< 1$ but converges to 1. Another: $a_n = 2 + 1/n \to 2$. The ratio condition guarantees monotonic decrease, and positivity provides a lower bound, ensuring convergence, but the specific limit depends on the sequence's infimum. This distinguishes guaranteed existence of a limit from the specific value of the limit, countering the misconception that 'decreasing positive' always means 'goes to zero'.

📖 Explanation: This Olympiad-style question explores multiple rigorous pathways to a non-trivial limit. Method A uses continuity of ln and transforms $1^\infty$ to $0 \cdot \infty$, solvable via L'Hôpital or Taylor series ($\ln(1-x) \approx -x$). Method B uses algebraic substitution to align with the standard definition $\lim (1-1/m)^m = e^{-1}$, noting $(1-1/m)^{-1} \to 1$. Both are mathematically sound and complementary. Recognizing multiple valid strategies demonstrates deep fluency with exponential limits. Direct substitution fails due to indeterminate form. This validates flexibility in analytical techniques for challenging sequence limits.

📖 Explanation: This piecewise-defined sequence explicitly constructs different behaviors for subsequences. Odd terms $\to 0$, even terms $\to 1$. Since $0 \neq 1$, the full sequence cannot converge (Theorem 9.1.4). It is bounded (between 0 and 2) but not monotone. This serves as a canonical example of divergence via subsequence mismatch. Students must recognize that convergence requires ALL subsequences to share the same limit. Dominance notions (options A/B) are irrelevant here; both subsequences are infinite. This tests precise application of subsequence criteria over intuitive but incorrect 'majority' reasoning.

📖 Explanation: The Squeezing Theorem requires bounds that converge to the SAME limit. Since $-1 \leq \cos(n) \leq 1$, dividing by $n^2$ gives $-1/n^2 \leq b_n \leq 1/n^2$. Both bounds $\to 0$, forcing $b_n \to 0$. Option A bounds don't converge to same limit. Option B bounds work but are looser than necessary (though still valid). Option D is invalid because $\cos(n)$ can be negative. Option C provides the tightest, most natural bounds derived directly from cosine's range, making it the most appropriate choice. This recalls the standard technique for trigonometric-over-polynomial sequences.

📖 Explanation: Convergence to 0 does not imply monotonicity. The sequence $(-1)^n / n$ converges to 0 (by Squeezing/Absolute Value Theorem), but it alternates signs: $-1, 1/2, -1/3, 1/4, ...$. This is not eventually decreasing because positive terms follow negative terms, violating monotonic decrease. Options A, C, D are all eventually decreasing. This counterexample reinforces that convergence concerns proximity to a limit, not directional consistency. Oscillatory decay is a valid convergence mode. Understanding this prevents erroneous assumptions in proofs where monotonicity is needed but not justified by convergence alone.

📖 Explanation: Biological models require both existence of equilibrium (fixed point $x=K$) and dynamic stability (monotone approach). Solving $x = Rx/(K+(R-1)x)$ gives equilibria 0 and K. Monotonicity (decreasing when $x>K$) ensures populations smoothly regulate toward carrying capacity K without chaotic oscillations or extinction. This combination validates the model as biologically realistic for regulated species. Pure fixed point analysis ignores dynamics; pure monotonicity ignores destination. Together they confirm stable equilibrium. This integrates mathematical rigor with domain-specific interpretation, showing why both theoretical tools are essential in applied modeling.

📖 Explanation: This tests understanding of continuity's role in limit evaluation. The step $\lim e^{g(n)} = e^{\lim g(n)}$ is valid ONLY because $e^x$ is continuous at the limit point (here, 0). If the outer function were discontinuous at the inner limit, this move would be invalid. This principle underpins many sequence limit evaluations involving transcendental functions. L'Hôpital justifies $\ln n / n \to 0$, not the composition. Monotonicity is irrelevant. Recognizing continuity as the enabler prevents misuse of this powerful technique in cases where functions have jumps or singularities at the limit point.

📖 Explanation: Rewriting as $(1/2)^n + (3/4)^n$ decomposes the problem into two geometric sequences with ratios $< 1$. Each term $\to 0$, so sum $\to 0$. This leverages known geometric limits directly. Factoring (B) works but is messier. L'Hôpital (A) is inappropriate for discrete exponentials without extension. Squeezing (D) is valid but unnecessary when exact decomposition exists. Choosing the simplest, most direct method demonstrates strategic competence. This highlights recognizing structure (sum of geometrics) over brute-force techniques, a key skill in efficient sequence analysis.

📖 Explanation: This Olympiad-level question invokes the Cauchy criterion. The condition implies $|a_m - a_n| \leq \sum_{k=n}^{m-1} |a_{k+1}-a_k| < \sum_{k=n}^{\infty} 1/2^k = 1/2^{n-1} \to 0$ as $n,m \to \infty$. Thus, $\{a_n\}$ is Cauchy. In complete metric spaces (like $\mathbb{R}$), every Cauchy sequence converges. This is stronger than mere term-to-term decay; it guarantees cumulative stability. Monotonicity isn't required. This tests knowledge beyond basic tests, connecting difference bounds to fundamental completeness properties, essential for advanced analysis.

📖 Explanation: Visual alignment with a continuous function having a horizontal asymptote directly implies sequence convergence to that asymptote value. Since $f(x) = x/(x+1) \to 1$ as $x \to \infty$, and $a_n = f(n)$, the sequence inherits this limit. The graph provides immediate intuitive confirmation of the function-sequence limit theorem. Points getting arbitrarily close to y=1 visually encode the epsilon-N definition. This reinforces translating graphical asymptotes into sequential limits, distinguishing from oscillatory or divergent patterns. Visual literacy complements analytical skills in understanding convergence.

📖 Explanation: Algebraic limit theorems (sum, product, etc.) operate on *existing* limits. They cannot create limits. The Monotone Convergence Theorem asserts existence where none was previously known. The Completeness Axiom (every bounded nonempty set has a supremum) provides this existential foundation: the limit IS the supremum of the sequence's range. Without Completeness (e.g., in rationals), bounded monotone sequences might lack limits (e.g., decimal approximations of $\sqrt{2}$). This distinguishes foundational axioms from operational rules, highlighting why real analysis requires more than algebra to establish convergence.

📖 Explanation: Substituting $x = 1/n$ transforms the sequence limit into the fundamental calculus limit $\lim_{x \to 0} \sin(x)/x = 1$. This converts indeterminate $\infty \cdot 0$ into a known standard form. Squeezing (A) is possible but less direct. Option C ignores the shrinking argument of sine. Option D misattributes oscillation; $\sin(1/n)$ doesn't oscillate as $n \to \infty$, it smoothly approaches 0. This demonstrates strategic variable substitution to leverage known continuous limits for discrete sequences, a vital technique for resolving indeterminate forms efficiently.

📖 Explanation: Series convergence necessitates $a_n \to 0$ (Divergence Test), but imposes NO monotonicity requirement. One could construct a convergent series with non-monotone terms (e.g., rearranged or modified). The student conflates a consequence of this specific formula's structure with a general consequence of convergence. While $1/n^2$ IS decreasing, it's not BECAUSE the series converges. This separates logical implication from coincidental property, correcting a subtle but important causal fallacy in relating series and sequence properties.

📖 Explanation: The recurrence is odd-symmetric: if $a_n < 0$, $a_{n+1} < 0$. Starting negative, it stays negative and converges to the negative fixed point $-\sqrt{2}$. This matters for robustness: algorithms must handle sign correctly or specify domain restrictions. Assuming positivity without enforcement risks wrong-sign convergence. This blends sequence dynamics with numerical method sensitivity, showing mathematical behavior depends critically on initial conditions and inherent symmetries, not just the formula's form.

📖 Explanation: $(-1)^n$ is bounded but diverges (oscillation). The Ratio Test gives $|a_{n+1}/a_n| = 1$, which is inconclusive. This dual failure makes it the perfect counterexample for both concepts. Other options either converge or are unbounded. This reinforces that boundedness ≠ convergence and that ratio=1 signals need for alternative methods. It's a foundational example linking multiple theoretical limitations in one simple sequence.

📖 Explanation: This connects discrete sequence limits to continuous physical laws. Recognizing $\lim_{n \to \infty} (1 - kT/n)^n = e^{-kT}$ shows the discrete drag model converges to exponential decay. This validates the numerical scheme as consistent with continuous physics. It's not average velocity or distance; it's the instantaneous state. This exemplifies how sequence limits bridge computational models and analytical solutions, crucial for verifying simulations against theory.

📖 Explanation: Limits preserve weak inequalities, not strict ones. Example: $a_n = 0$, $b_n = 1/n$; $0 < 1/n$ but limits are both 0, so $L = M$. Thus, only $L \leq M$ is guaranteed. Strict inequality can collapse in the limit. This is a fundamental property often misunderstood. Recalling this prevents erroneous strict conclusions in limit comparisons, emphasizing the topological nature of limits where boundaries may touch.

📖 Explanation: Ratio is simplest here: $\frac{(n+1)!}{e^{n+1}} \cdot \frac{e^n}{n!} = \frac{n+1}{e} \to \infty$. Clearly >1, so $a_n \to \infty$. Root test involves $\sqrt[n]{n!}$ which requires Stirling or prior knowledge. Integral test inapplicable (discrete factorial). Comparison possible but ratio is direct. This showcases selecting the optimal test based on term structure (factorial/exponential mix favors ratio), demonstrating strategic efficiency in convergence analysis.

📖 Explanation: This question targets error analysis regarding the definition of convergence. Students often conflate oscillation with divergence. While the sequence indeed diverges, the justification 'alternating signs' is flawed because sequences like $(-1)^n/n$ alternate yet converge to zero. The correct analysis requires evaluating $\lim_{n \to \infty} a_n$. Since $\lim_{n \to \infty} |a_n| = 1 \neq 0$, the sequence fails the necessary condition for convergence, making the conclusion right but the reasoning fundamentally unsound.

📖 Explanation: This addresses the subtle distinction between sequential limits and functional limits. While $f(n) = \sin(n\pi) = 0$ for all integers, causing the sequence to converge trivially to 0, the continuous function $\sin(\pi x)$ oscillates perpetually between -1 and 1 as $x \to \infty$. This demonstrates that convergence of a sequence $\{f(n)\}$ does not imply convergence of the underlying function $f(x)$, a critical conceptual nuance in understanding sequences as discrete samplings rather than continuous behaviors.

📖 Explanation: This applies the formal epsilon-N definition to a modeling scenario. In chaotic regimes of logistic maps, trajectories are bounded but aperiodic. Formally, a limit L requires that for every $\epsilon > 0$, there exists N such that $|a_n - L| < \epsilon$ for all $n > N$. Chaotic behavior violates this because terms continue to visit distant regions of the state space indefinitely. Thus, despite being generated by a deterministic rule, the sequence lacks a limit, illustrating that 'having a formula' does not guarantee convergence.

📖 Explanation: This tests graph-based interpretation of sequence definitions. Sequences are functions with domain $\mathbb{N}$, so their graphs must consist of discrete, isolated points, eliminating continuous curve options. For this specific sequence, the points should show a strictly increasing trend that flattens out, suggesting a horizontal asymptote (the number e). Recognizing the discrete nature of sequence graphs versus continuous function graphs is fundamental, and visually identifying monotonic boundedness provides intuitive evidence of convergence consistent with the Monotone Convergence Theorem.

📖 Explanation: This covers direct recall of the relationship between functional and sequential limits. A core theorem states that if a function $f(x)$ approaches a limit L as $x \to \infty$, then the sequence formed by sampling at integers, $f(n)$, must also approach L. This is because the integer inputs form a subset of the real domain. However, the converse is false. Understanding this directional implication is foundational for applying calculus tools like L'Hôpital's Rule to sequence problems by extending them to continuous domains.

📖 Explanation: This error analysis question highlights the domain restriction in sequence definitions. Sequences are discrete; derivatives require continuity. One cannot differentiate $n!$ with respect to n directly. To use L'Hôpital's Rule, one must extend the sequence to a continuous function, typically using the Gamma function $\Gamma(x+1)$ for factorials, or use alternative methods like the Squeezing Theorem or ratio test logic. Direct application of differential calculus to discrete indices is a common misconception stemming from ignoring the fundamental definition of a sequence as a function on integers.

📖 Explanation: This challenges conceptual understanding by testing necessary vs. sufficient conditions. While convergent sequences must have differences approaching zero, the converse fails. For $a_n = \ln n$, the difference $\ln(n+1) - \ln(n) = \ln(1 + 1/n) \to 0$, yet $\ln n \to \infty$. This distinguishes 'terms getting closer together' from 'terms approaching a finite limit.' Students often confuse the Cauchy criterion (which involves arbitrary gaps, not just adjacent ones) with simple adjacent differences. This example reinforces that vanishing increments do not guarantee boundedness or convergence.

📖 Explanation: This application question connects physical measurement constraints to mathematical definitions. Real-world sensors produce discrete data points, naturally forming a sequence. While the underlying physical law is continuous, the dataset is a sequence. This distinction matters because statistical tools for sequences differ from those for continuous functions, and interpolation assumptions introduce error. Mathematically, treating data as a sequence respects the domain of observation. Convergence questions about 'eventual safety levels' become questions about sequential limits, emphasizing that the mathematical model must match the discrete nature of empirical evidence collection.

📖 Explanation: This tests conceptual understanding of the Monotone Convergence Theorem as it relates to sequence definitions. Recursive definitions don't automatically yield explicit formulas. To assert a limit exists without finding it, one relies on structural properties. Showing $a_n < 2$ (boundedness) and $a_{n+1} > a_n$ (monotonicity) invokes the completeness axiom of real numbers, guaranteeing a supremum that serves as the limit. This shifts focus from computation to existence proofs, highlighting that the definition of convergence is deeply tied to the order structure of real numbers, not just algebraic manipulation.

📖 Explanation: This addresses a pervasive misconception: equating boundedness with convergence. While every convergent sequence is bounded, the reverse isn't true. $\sin(n)$ stays within [-1, 1] but never settles; the integer arguments modulo $2\pi$ are dense in the circle, causing perpetual erratic oscillation. This contrasts with $\sin(1/n)$, which converges to 0. The explanation reinforces that the definition of limit requires eventual proximity to a specific L, not merely confinement within a range. Boundedness prevents divergence to infinity but permits non-convergent oscillation.

📖 Explanation: This is a direct application of standard techniques derived from limit definitions. Dividing by the highest power of n transforms the expression into a form where each term's limit is obvious via basic arithmetic rules ($1/n^k \to 0$). While L'Hôpital's works after continuous extension, algebraic simplification is more fundamental to sequence theory as it relies purely on sequential limit laws without invoking derivatives. This method directly operationalizes the definition by reducing complex ratios to sums of known null sequences, reinforcing the hierarchy of growth rates inherent in polynomial sequences.

📖 Explanation: This mixed-concept question links sequence limits to famous continuous limits via substitution. Letting $x = 1/n$, as $n \to \infty$, $x \to 0^+$. The expression becomes $\frac{\sin x}{x}$, whose limit is 1. This illustrates how sequence problems often reduce to standard calculus limits through variable change. It tests recognition of structural equivalence between discrete and continuous forms. Understanding this connection validates using continuous calculus tools for sequences while respecting domain transformations, bridging the gap between Chapter 9 sequences and earlier limit concepts.

📖 Explanation: This error analysis targets the preservation of inequalities under limits. If $a_n < M$ for all n, then $\lim a_n \leq M$, not necessarily $< M$. Strict inequalities can become equalities in the limit (e.g., $1 - 1/n < 1$ but limit is 1). This subtle point arises directly from the epsilon-definition: for any $\epsilon$, terms are within $\epsilon$ of L, allowing L to touch the boundary. Failing to grasp this leads to incorrect bounds estimation and misunderstands the topological nature of limits as closure operations rather than pointwise constraints.

📖 Explanation: This application question tests understanding of 'eventual' properties versus global ones. Stock prices often exhibit initial volatility (non-monotone) before trending (monotone). Mathematical definitions accommodate this: a sequence is eventually monotone if discarding finitely many initial terms yields monotonicity. Convergence depends only on tail behavior, making initial irregularities irrelevant. This contrasts with idealized physics models (decay, interest) that are monotone from start. Recognizing real-world data as 'eventually' well-behaved validates applying convergence theorems to messy empirical datasets, emphasizing robustness of limit definitions against transient anomalies.

📖 Explanation: This tests strategic selection of methods based on form analysis. Naive evaluation suggests difference of infinities, providing no information. Rationalizing yields $\frac{1}{\sqrt{n+1}+\sqrt{n}}$, clearly showing convergence to 0. This highlights that sequence definitions involve limiting processes, not arithmetic evaluation. Indeterminate forms signal hidden structure requiring algebraic revelation. Choosing appropriate transformations is a higher-order skill beyond rote procedure; it requires diagnosing why direct methods fail and selecting tools that expose asymptotic behavior, reinforcing that limits describe trends, not static values.

📖 Explanation: This mixed-concept question probes algebraic properties of limits. If sum converged, then $b_n = (a_n + b_n) - a_n$ would be difference of convergent sequences, hence convergent—contradiction. Thus, adding a convergent sequence to a divergent one preserves divergence. This tests logical deduction from definitions rather than memorization. Students might incorrectly think 'convergent + divergent = indeterminate,' but linearity of limits makes this determinate. Understanding these interaction rules prevents errors in decomposing complex sequences and reinforces that divergence is a persistent property under translation by convergent perturbations.

📖 Explanation: This foundational conceptual question distinguishes sequences from series. Students frequently confuse $\lim a_n$ with $\sum a_n$. The sum is defined as $\lim s_n$ where $s_n = \sum_{k=1}^n a_k$. Even if $a_n \to 0$ (necessary), the accumulated sum may diverge (harmonic series). This distinction is paramount: sequences list values; series aggregate them. Confusing these leads to catastrophic errors like claiming harmonic series converges because terms vanish. Reinforcing this definitional separation is the first step in mastering infinite series.

📖 Explanation: This Olympiad-style question connects numerical analysis rates to sequence convergence definitions. Quadratic convergence means error squares each iteration, far exceeding geometric (linear) convergence. While standard definitions only ask 'does it converge?', rate analysis quantifies 'how fast'. This matters practically: quadratic sequences reach machine precision in few steps. Understanding this enriches the abstract definition with computational reality. It shows that while all convergent sequences share the same limit definition, their practical utility varies enormously based on error decay dynamics, linking pure analysis to algorithmic efficiency.

📖 Explanation: This graph-based question tests visual recognition of sequence properties. Key features: discreteness (not continuous), alternation (due to $(-1)^n$), and decay envelope $1/n$. Option B fails because sequences aren't continuous curves. Option C misses sign alternation. Option D ignores nonzero terms. Correct identification requires synthesizing multiple attributes into a visual representation. This reinforces that sequence graphs are point sets in $\mathbb{N} \times \mathbb{R}$, and visual intuition about damping envelopes supports analytical verification of convergence to zero despite oscillation.

📖 Explanation: This direct recall question links calculus derivatives to sequence monotonicity. If the continuous extension has negative derivative, the function decreases everywhere, implying $f(n+1) < f(n)$. Thus the sequence is strictly decreasing. Note: this doesn't guarantee convergence (could go to $-\infty$) nor positivity. Students sometimes assume decreasing implies convergence to zero, but monotonicity alone only gives direction, not destination. This question isolates the monotonicity inference from convergence conclusions, ensuring precise understanding of what derivative information actually provides about discrete sequences.

📖 Explanation: This mixed-concept question explores parametric sequences. This specific sequence is the constructive definition of $e^x$. Varying x changes the limit continuously, demonstrating how sequences can generate functions. This bridges sequences and transcendental functions, showing limits aren't just numbers but can define entire mathematical objects. Understanding this elevates sequences from computational exercises to foundational definitions in analysis. It reveals that 'limit' is a dynamic operator mapping parameters to values, central to defining exponentials rigorously without circular reliance on prior exponential definitions.

📖 Explanation: This application question validates Squeezing Theorem usage for oscillatory sequences. Cosine's irregularity at integers makes direct limit evaluation impossible, but boundedness combined with decaying envelope suffices. The theorem's power lies in bypassing internal complexity via external control. Students must recognize when detailed behavior is irrelevant compared to dominating trends. This exemplifies how convergence definitions handle 'messy' sequences: we don't need to track every fluctuation, only establish containment within shrinking corridors. This strategic ignorance is key to analyzing complex sequences efficiently.

📖 Explanation: This direct recall tests precise epsilon-N definition comprehension. Convergence means eventual permanent closeness, not universal closeness (early terms can be far) nor exact equality (asymptotic approach). Also, the number of outliers isn't specified, only finiteness. Option A captures the existential quantifier and universal tail condition correctly. Misunderstanding this leads to errors in proofs and misinterpretation of approximation accuracy. Mastery of this logical structure is prerequisite for all rigorous analysis, distinguishing vague 'getting close' from precise mathematical convergence.

📖 Explanation: This conceptual question dissects divergence mechanisms. Boundedness + subsequence convergence ≠ convergence. The definition requires ALL tails to be near ONE L. Here, subsequences converge to ±1, violating uniqueness. This illustrates that convergence demands coherence across the entire sequence, not just parts. Students often mistakenly believe bounded oscillation implies some averaged limit, but standard convergence is strict. Recognizing subsequence disagreement as definitive divergence proof is a powerful diagnostic tool, especially for alternating or piecewise-defined sequences where direct epsilon arguments are cumbersome.

📖 Explanation: This challenging question merges numerical reality with theoretical definitions. Pure math assumes exact arithmetic; computers don't. Persistent roundoff creates an error band preventing true convergence to L. Observed sequence stabilizes within $\pm 10^{-6}$ of L but may never enter smaller epsilon-bands. This reveals tension between idealized definitions and practical computation. Students must understand that numerical 'convergence' is approximate, and stopping criteria must account for machine precision. This bridges pure analysis and numerical methods, showing definitions need adaptation for real-world implementation where exact limits are physically unrealizable.

📖 Explanation: This application question tests sensitivity of limits to algebraic structure. Original limit is 1 (same degree). Changing numerator to constant creates degree mismatch, yielding 0. Other options preserve degree balance or merely alter sign/convergence status without changing magnitude limit. This reinforces that polynomial sequence limits depend on leading term ratios. Students learn to predict limit changes structurally rather than recomputing each time. Developing this intuition accelerates problem-solving and deepens understanding of asymptotic dominance hierarchies embedded in sequence definitions.

📖 Explanation: This Olympiad-style question stress-tests Squeezing Theorem conditions. The theorem requires inequality holding eventually for ALL n, not just a subsequence. Even-index squeezing controls only half the sequence; odd terms remain unconstrained and could diverge. This exposes fragility of convergence: controlling a subset doesn't control the whole. Students must verify hypotheses completely, not partially. This cultivates rigor in checking quantifiers ('for all' vs 'for some') and prevents overgeneralization of powerful theorems beyond their valid domains.

📖 Explanation: This mixed-concept question clarifies categorical distinctions. Both use epsilon language but in different contexts: sequences → ∞ discretely; continuity → c continuously. Confusing them leads to category errors like discussing 'continuity at infinity' or 'sequential continuity' without proper framing. Understanding this separation helps organize analysis mentally: sequences capture asymptotic discrete trends; continuity captures local smoothness. Though related (sequential criterion for continuity), they address different mathematical phenomena. Articulating this difference strengthens conceptual architecture of calculus.

📖 Explanation: This challenging question targets the gap between local and global behavior. Vanishing adjacent differences suggest slowing change but don't prevent unbounded drift (e.g., $\sum 1/n$). True Cauchy condition requires $|a_m - a_n| < \epsilon$ for ALL large m,n, not just neighbors. This distinction is subtle but crucial: local smoothness ≠ global stability. Students often assume 'slowing down' means 'stopping,' but cumulative small changes can still yield divergence. This deepens understanding of what convergence truly demands beyond superficial trend inspection.

📖 Explanation: This application question emphasizes domain fidelity in modeling. Financial contracts specify discrete compounding dates; interpolating continuously introduces fictitious values. While continuous approximation aids intuition, exact answers require respecting discrete sequence definition. This teaches that mathematical convenience shouldn't override problem constraints. Students learn to choose representations matching reality: sequences for discrete events, functions for continuous flows. Misalignment causes errors in timing-sensitive applications. This reinforces that definitions carry semantic meaning beyond syntax, anchoring math to the phenomena it describes.

📖 Explanation: This tests multi-step reasoning combining extension and calculus. Direct sequence L'Hôpital is invalid; must extend to continuous f(x), compute limit, then transfer back via theorem. Option B's bound is too weak (<1 doesn't imply →0). Option C is false. Option D's upper bound 1/n is incorrect (ln n / n > 1/n for large n). Only A follows valid logical chain: extend → differentiate → conclude → restrict. This workflow exemplifies standard technique for indeterminate sequence forms, reinforcing proper protocol for leveraging continuous tools on discrete objects.

📖 Explanation: This conceptual question reinforces uniqueness of limits. Having convergent subsequences isn't enough; they must converge to SAME value. Here, even→0, odd→1, so no unique attractor exists. This violates the 'for every epsilon' clause: pick ε=0.4, and no N keeps all subsequent terms within 0.4 of any candidate L. Students sometimes think 'settling into pattern' equals convergence, but periodic patterns ≠ limits. This clarifies that convergence demands asymptotic constancy, not just regularity, sharpening intuition about what 'approaching a value' truly means.